Turbo/ Super Charged Torque and Power Curves

[oz_olly]

Hey

I am doing my thesis on transmission selection in formula SAE and as such I would like to get my hands on some forced induction torque and power curves. My proposal is as follows - the ideal transmission would allow you to operate at constant power so for a peakyish naturally aspirated torque curve you need infinite gear ratios to achieve this as the rpm range of peak power is so small i.e. a CVT. However I feel that with a turbo sort of torque curve you can more or less operate at constant power through careful selection of gear ratios and possibly few gears as FSAE doesn't go so fast.

So to get to the point I would be extremely greatful if anyone from a forced induction team would be willing to send me torque and power curves in the tabulated form i.e. an excel spread sheet, although a plot would also be good. I have seen the plots on this page but if i could get tabulated data it would be the big wicked. Anyone who helps be out will be fully acknowleged and I will even devote an entire paragraph explaining all the kudos that this individual will have brought upon themself.

If willing to help please email to m.olsen@student.adfa.edu.au

Cheers

Olly

Team Leader
ACME Racing
UNSW@ADFA

[oz_olly]

I would like to publicly extend my sincere thanks to Mr Michael Royce for being kind enough and interested enough to forward my request to William Wattard. For those who don't know he designed and built his own parallel two cylinder engine for FSAE, he has written several papers on the work he and his research partners have completed mainly concentrating on the benefits of forced induction. His work is of an extremely high standard and I have been genuinely inspired by his work ethic and dedication. Big thanks to William and congratulations on the awesome work.

Cheers

Olly

Team Leader
ACME Racing
UNSW@ADFA

P.S. Anyone else willing to share information will be extremely appreciated.

[Welfares]

He's William Attard actually, he nicknamed his engine the "Wattard" engine.

He used to post, don't know if he's still on the forum. He was at the event this year with Melbourne. Good Bloke, Smart man.

[oz_olly]

Oops my bad, it was a typo. Apologies to William.

[Homemade WRX]

now I wan't to know more.

[rjwoods77]

http://www.mame.mu.oz.au/formulasae/SAE%20paper%202006-01-0745.pdf

[JerryLH3]

William is an incredibly nice guy. I met him at the Motorsports Engineering conference in December. He was extremely helpful in answereing any questions I came up with and providing info.

The development work done on the WATTARD engine is unbelievable. He's spent lots of time on it and there are a few SAE papers detailing it. I imagine he would be more than willing to help out.

[powered by wattard]

Thanks guys. Its been a while since I posted anything, busy with the Phd.

Thanks also to all th guys I met at the Small Engine Confernece in San Antonio and the Motorsports Conference in Detroit for their hispitality, being so far from the land down under (Melbourne Australia). Here are the titles of some SAE technical papers I have written in the last year or so. The PDF's can all be found by googling the paper number and viewing through eprints or at (http://www.mame.mu.oz.au/formulasae/) and following the link to the Research page. Titles are below and are all extremly relevant to FSAE engine guys no matter if NA or TC . Note the first paper solves the oil control issues for throttled compressors for all you turbocharging guys who are having oil issues. It is sheduled to be presented at the world Congress in April 2007 in Detroit but is not on the web yet. I will post it in the next few days.

2007-01-1562
Highly Turbocharging a Flow Restricted Two Cylinder Small Engine - Turbocharger Development

2006-01-3637
http://eprints.infodiv.unimelb.edu.au/archive/00002551/01/2006-01-3637.pdf
Highly Turbocharging a Restricted, Odd Fire, Two Cylinder Small Engine - Design, Lubrication, Tuning and Control

2006-01-0745
http://eprints.infodiv.unimelb.edu.au/archive/00002725/...per_2006-01-0745.pdf (http://eprints.infodiv.unimelb.edu.au/archive/00002725/01/SAE_paper_2006-01-0745.pdf)
Development of a 430cc Constant Power Engine for Formula SAE Competition

2006-32-0072
http://eprints.infodiv.unimelb.edu.au/archive/00002548/01/2006-32-0072.pdf
Comparing the Performance and Limitations of a Downsized Formula SAE Engine in Normally Aspirated, Supercharged and Turbocharged Modes

2006-32-0036
http://eprints.infodiv.unimelb.edu.au/archive/00002550/...06-32-0036colour.pdf (http://eprints.infodiv.unimelb.edu.au/archive/00002550/01/2006-32-0036colour.pdf)
Design and Development of a Gasketless Cylinder Head / Block Interface for an Open Deck, Multi Cylinder, Highly Turbocharged Small Engine

Regards
William Attard
University of Melbourne

PS: Thanks to all you guys from Formula SAE teams that I met while I was in the US. Its always good to share stories and all FSAE people know what it is all about. FSAE doods are cool, no matter what part of the world you are from

[absolutepressure]

Originally posted by Rob Woods:
http://www.mame.mu.oz.au/formulasae/SAE%20paper%202006-01-0745.pdf

Trying to interpret those charts gave me a headache. I'm guessing the dashed lines are torque curves. At first I thought all the lines were horsepower AND torque curves, as I have seen in some magazines, which confused me as to why there were two sets of lines, and why you were making 240 hp with 90Nm at 6000rpm. But I think i figured it out.

I noticed that when the engine isn't limited, you have a fairly large negative jerk* (the derivative of accel) you constantly accelerate at a decreasing rate, which would feel to me like the engine is dying down as rpm's increase (just a guess). I guess what I'm trying to say is that even though the torque was always higher than the other engines, it would feel like the engine was giving up on me, and that's not very inspiring. Again, I might be wrong, but that's just what I'm getting from the charts.

The second thing I noticed was that you said you were cranking out just under 2 atm's from the turbo. That's what, about 25psi? That's a LOT. Supras get that from a big T66. How do you get that out of a GT12?

Lastly, why were you shooting for a constant power output as opposed to a constant torque output? Horsepower is just a function of torque, so yes, to get constant power as rpm increases, you need to decrease torque output (as you have shown, severely), the thing that acually moves the car. Why did you have to limit the output in the first place? Was it really difficult to drive? Just because you have a lot of torque doesn't mean you have to use all of it all of the time. Did the drivers just like their foot hitting the positive stop?

I think it's awesome how you designed and built your own engine, but some things just don't make sense to me.
* jerk doesn't need to be violent when refering to it as the derivative of acceleration.

[Beatle]

To echo what many people have already said, I met William in Detroit at the Motorsports Conference. His presentation was excellent and I couldnt believe how many engine parameters he tested. It was definitely the best presentation given at the conference, although Milliken's presentation on tire testing for RC cars was pretty cool too.

The reason why Supras need T66's to get 25 psi of boost is because they are consuming huge amounts of air (big engine, big flow rate). A 430cc doesnt create such a large flow rate and therefore can use a smaller turbo. If you look at the GT12 comp. map you will see that the pressure ratio can reach over 3 (2 bar of boost) at around 9-10 lbs/min.

I think the reason he was trying to achieve constant power (and William if I am wrong please correct me) was because he wanted to have choked flow across the entire operating range (6k-10k). Turbocharged, restricted motors inherently have a decreasing torque curve becuase they are choking the restrictor, meaning they have the same flow rate at increasing engine speeds, resulting in nearly constant power, but decreasing torque. And Im pretty sure that if your engine is putting out 65 lb-ft of torque, and then slowly drops to 50 lb-ft, you wont be thinking that the engine is giving up on you.

Where does he say he is making 240 hp? The most I saw was around 80 hp?

[absolutepressure]

Originally posted by Beatle:
The reason why Supras need T66's to get 25 psi of boost is because they are consuming huge amounts of air (big engine, big flow rate).
Ok, that's kind of what I was thinking, but I wasn't sure.


Turbocharged, restricted motors inherently have a decreasing torque curve becuase they are choking the restrictor, meaning they have the same flow rate at increasing engine speeds, resulting in nearly constant power, but decreasing torque.

Oh, so it's not so much that he was shooting for constant power as opposed to torque. It's just that he was settling for the best he could get.


And Im pretty sure that if your engine is putting out 65 lb-ft of torque, and then slowly drops to 50 lb-ft, you wont be thinking that the engine is giving up on you.

Is that the conversion? I tried using the units converter on my calculator, but it didn't have torque. I saw a drop of 40 units and a spike that resembles Everest and went "Wow...that's....different." What are the actual conversion factors going both ways? So yeah, if it's only 15lb-ft then that's not as bad as it looks.


Where does he say he is making 240 hp? The most I saw was around 80 hp?
When I was trying to understand the graph I saw that the spike lined up with both 180kW and 90Nm at 6000rpm. I can convert power on my calculator and realized that I was reading something wrong. Pretty much all my confusion came from the graph layouts, their lack of a sufficient legend, and the inability to flip flop from SI to crap...I mean English.

I think I'm beginning to understand, but William, if you could chime in I think things would become clear to me.

[Beatle]

The torque does drop below 50 lb-ft, Im just saying that even though it is decreasing it is still making way more torque than most engines out there. From what I saw the max torque was around 67 lb-ft at about 6000 RPM, and it drops to 40 lb-ft by 10,000RPM, certainly a drop, but your still pulling pretty hard.

[bigwig]

powered by wattard, I've read your first article and I think a lot was missed in regards to your exhaust manifold design. Comparing the log manifold with a huge volume to that dinky manifold with actual runners. It clearly shows in your data, that due to the larger volume of the exhaust manifold you were able to increase your knock limit due to a decrease in exhaust reversion. I would almost gaurentee that using a manifold with a nice merge collector and decent primary length you would have come to a completely different result, especially in 9000-11000rpm range.

If you guys into turbos want to be fancy, I suggest fancy dancy boost controllers. Controlled boost creep is just about the best way to get constant power/torque curves.

One more comment, the difference in big turbos and small turbos is not only the compressor but the turbine. Ideal gas law tells us that pressure can only increase when volume and temperature changes. With a decent intercooler, temperature becomes less of a key point so the next thing to look at is volume. Well the volume includes the engine obviously. When you have a dinky turbine and high exhaust manifold pressure to intake manifold pressure ratios, tons of reversion occurs. This means to get that constant pressure, the turbo needs to flow less air because the given volume is being taken up by old gas. When you have big turbos, with big turbines, you have less reversion, thus the compressor can keep flowing air and power increases. People realize this more and more now and thats why most turbocharged systems are basically being built like NA setups. They attempt to run 1:1 pressure ratios and run pretty aggressive NA cams, use huge compressors which flow tons of air. Your VE skyrockets and you make tons of power.

PS: I'm too lazy to proof read any of this so some clarification will probably be needed.

[repeatoffender]

pretty big talk for someone thats obviously never tried to implement a turbo system on a restricted small capacity engine before bigwig.

so what your telling him is to increase runner length, and inertia of the compressor - turbine assembly to improve power.

have you thought about what happens to the system response because of this?

considering how on off throttle that fsae is how can you suggest this as a good idea?

even better, whos to say that all the extra power that can be gained will be used?

ive talked to a few teams that have turbo setups (ie UTA,Cornell, WOllongong) the main advantage they all have over NA setups of similar power (ie ~ 100 HP) is a wide, yes wide and large torque band.

[bigwig]

Would you care to look at some turbo F1 setups? Did they use log manifolds like that? I don't think so. I was just making note that the fact he compared two manifolds on unequal volume it is not the best test. Look at the power changes from the different volumes in log manifolds. You don't think the same applies for tubed manifolds? http://www.full-race.com/dyno_testing.php Here there are 3 tests on 2.0L or smaller engines comparing small volume log manifolds and long runner equal length manifolds. The EL manifolds make gobs more power everywhere. Again, it would have just been nice to see a tubed manifold with some runner length and maybe it would illustrate something worth while as current turbo theory is to use manifolds with runner length. But what the fuck do I know?

If you want to know what I'm suggesting. I'm suggesting people look into longer runnered, equal length exhaust manifolds with nice merge collectors. Also use turbos with slightly smaller turbines and slightly larger turbines and use the wastegate to regulate exhaust manifold pressure rather than the turbine. Turbine is just a slave to the compressor so the faster you get the compressor spinning the better, and smaller turbines generally spin easier than bigger ones. I also suggest if you want torque to remain constant, a controlled boost creep is the best way of doing so, especially in the occassion. Compressor maps generally allow more flow at higher pressures so yea, good way to keep torque up. But again, I'm a big talker. I thoroughly appologize for pointing out something I felt was neglected.

[repeatoffender]

Originally posted by bigwig:
Would you care to look at some turbo F1 setups? Did they use log manifolds like that? I don't think so. I was just making note that the fact he compared two manifolds on unequal volume it is not the best test. Look at the power changes from the different volumes in log manifolds. You don't think the same applies for tubed manifolds? http://www.full-race.com/dyno_testing.php Here there are 3 tests on 2.0L or smaller engines comparing small volume log manifolds and long runner equal length manifolds. The EL manifolds make gobs more power everywhere. Again, it would have just been nice to see a tubed manifold with some runner length and maybe it would illustrate something worth while as current turbo theory is to use manifolds with runner length. But what the fuck do I know?

If you want to know what I'm suggesting. I'm suggesting people look into longer runnered, equal length exhaust manifolds with nice merge collectors. Also use turbos with slightly smaller turbines and slightly larger turbines and use the wastegate to regulate exhaust manifold pressure rather than the turbine. Turbine is just a slave to the compressor so the faster you get the compressor spinning the better, and smaller turbines generally spin easier than bigger ones. I also suggest if you want torque to remain constant, a controlled boost creep is the best way of doing so, especially in the occassion. Compressor maps generally allow more flow at higher pressures so yea, good way to keep torque up. But again, I'm a big talker. I thoroughly appologize for pointing out something I felt was neglected.

Again i fail to see your point and how this applies to FSAE conditions.

If Formula 1 was to jump off a bridge would you follow? Yes F1 engines were restricted, however they did not have the throttle before the compressor nor were they restricted to less than 610cc engines!

First off you suggest big turbos with larger turbines, now you are suggesting smaller turbines and larger compressors. Which is it?

Maybe he was driven by packaging rather than straight performance for the exhaust as it exited over the gear box. However a great merge collector before the turbine housing really does help. The flow characteristics through the plenum in the exhaust may have been sufficient in his design. From the pictures i saw, his runners were equal length. How about this, even though its not used in a turbocharged application, MotoGP 4stroke engines headers have similar 'powerband' expansion chambers as to those found in 2stroke bikes!

I should have defined the term small capacity engines for you as being less than 1 litre!

If your so into magic torque generation through boost creep control, how about changing to the GT15 and using the vane type control method?

What sort of control method are you talking of? modifying the pressure signal? or electrically actuated wastegate? Tell me where are you going to get a reliable pressure signal from now you have the throttle before the compressor?

So instead of asking someone else to do all the work for you, you do the testing, you write up the results, and then WE will critisize you for not testing enough setups.

[powered by wattard]

Bigwig, Dont forget to look at the simulation data for varying exhaust geometries in the paper 2006-01-3637. The engine is very different to many out there. It is odd fire (0-180) due to the inline twin meaning the turbine recieves unequal pulses thus effecting turbine velocities, spool rates and achievbable boost. Thus a wider power curve can be achieved with constant press over pulse type trubocharging with equal exhaust volumes due to dampening the unequal pulsations. This is very different to any 4 cylinder or old F1 engines. If equal fire, previous published results from many sources agree with your statements, as do I about exhaust lenghts with a nice collector. I should send you the chapter from the thesis, has much more data but you can only put so much in the paper. The performance reduction you see from the pulse system is only partly attributed to the residuals, with dilivered boost also varying between the 2 systems for several reasons (knock and achievable boost). Longer equal pipes in pulse equal fire setups also mean more heat losses and poor drivability. We also tried unequal lenght pipes in the exhaust system for the pulse setup to ensure equal pulses to the turbine, but this is speed dependant and thus could only be tuned to eliminate odd pulsing at a given engine speed.

[Chuck Dean]

If you want to know what I'm suggesting. I'm suggesting people look into longer runnered, equal length exhaust manifolds with nice merge collectors.

I don't agree with you 100% on this. Equal length is good. But there is often a happy medium between super short runners and long TUNED runners. By keeping the runners short you take advantage of the high energy exhaust. I'm not saying that what you said is wrong either. There are pro and cons, performance gains associated with both types of manifolds. If you are able to get better performance out of the long tuned runners, it is often very small and not worth the weight/packaging problems. Unless of course the the long runner packaging is advantageous to your particular application. You can achieve NEAR equal length runners of short length and still get them to collect nicely into the turbine.

I have an SAE paper that backs this up. They ran a simulation that optimized the length. I have also done some case studies WAVE as well.

And again, I'm not speaking on William's particular case. I'm responding to your suggestion that LONG tuned equal length is the way to go. It's not as simple as that.

[Chuck Dean]

Also, the reason why the full race manifold showed such high gains is because it's being compared to a poorly designed log manifold. The gains are more than likely due to the fact the exhaust was actually collected properly before being dumped in the turbine, not due to the runner length. And those aren't what I would call LONG runners either.

(talking about the full race B18 test)

[Beatle]

Bigwig, you have a wonderful way of sounding correct even though what you are saying is pretty shortsighted.

As William said the larger volume of the log manifold will decrease the odd pulses from the engine and therefore prevent varying turbo performance, but like you said, what the fuck do you know.

As for the long exhaust runners, they may of worked well on an F1 car, but F1 cars dont run the same types of courses as FSAE cars. We run much slower courses with many more turns, and therefore some people would consider response a design parameter. If you have a long exhaust manifold you have more volume to pressurize and longer "reaction" times. This doesnt mean compensations cant be made (mainly in the size of the turbine), but if I were designing a turbo system, minimizing both the volume of both intake and exhaust would be my first step.